Access the full text.
Sign up today, get DeepDyve free for 14 days.
G. Mumtaz, S. Awad, A. Feizzadeh, H. Weiss, L. Abu-Raddad (2018)
HIV incidence among people who inject drugs in the Middle East and North Africa: mathematical modelling analysisJournal of the International AIDS Society, 21
G. Djomand, S. Quaye, P. Sullivan (2014)
HIV epidemic among key populations in west AfricaCurrent Opinion in HIV and AIDS, 9
T. Rhodes, Andy Guise, J. Ndimbii, S. Strathdee, E. Ngugi, L. Platt, A. Kurth, C. Cleland, P. Vickerman (2015)
Is the promise of methadone Kenya’s solution to managing HIV and addiction? A mixed-method mathematical modelling and qualitative studyBMJ Open, 5
S. Weir, S. Baral, Jessie Edwards, S. Zadrozny, J. Hargreaves, Jinkou Zhao, K. Sabin (2018)
Opportunities for Enhanced Strategic Use of Surveys, Medical Records, and Program Data for HIV Surveillance of Key Populations: Scoping ReviewJMIR Public Health and Surveillance, 4
R. Steen, J. Hontelez, Andra Veraart, R. White, S. Vlas (2014)
Looking upstream to prevent HIV transmission: can interventions with sex workers alter the course of HIV epidemics in Africa as they did in Asia?AIDS, 28
G. Meyer-Rath, J. McGillen, D. Cuadros, T. Hallett, S. Bhatt, N. Wabiri, F. Tanser, T. Rehle (2018)
Targeting the right interventions to the right people and places: the role of geospatial analysis in HIV program planningAIDS (London, England), 32
Highlights important lessons learned in doing optimization with dynamic models including all key populations in a variety of epidemic settings
C. Beyrer, S. Baral, C. Collins, E. Richardson, P. Sullivan, Jorge Sanchez, G. Trapence, E. Katabira, Michel Kazatchkine, Owen Ryan, A. Wirtz, K. Mayer (2016)
The global response to HIV in men who have sex with menThe Lancet, 388
S. Baral, Rachael Turner, C. Lyons, S. Howell, B. Honermann, A. Garner, Robert Hess, D. Diouf, G. Ayala, P. Sullivan, G. Millett (2018)
Population Size Estimation of Gay and Bisexual Men and Other Men Who Have Sex With Men Using Social Media-Based PlatformsJMIR Public Health and Surveillance, 4
Y. Caro-Vega, C. Rio, V. Lima, B. Crabtree-Ramírez, S. Bautista-Arredondo, C. Rio, M. López-Cervantes, M. Colchero, J. Sierra-Madero (2015)
Estimating the Impact of Earlier ART Initiation and Increased Testing Coverage on HIV Transmission among Men Who Have Sex with Men in Mexico using a Mathematical ModelPLoS ONE, 10
T. Saidel, D. Jarlais, W. Peerapatanapokin, J. Dorabjee, Siddharth Singh, T. Brown (2003)
Potential impact of HIV among IDUs on heterosexual transmission in Asian settings: scenarios from the Asian Epidemic ModelInternational Journal of Drug Policy, 14
C. Mukandavire, Josephine Walker, S. Schwartz, M. Boily, L. Danon, C. Lyons, D. Diouf, B. Liestman, Nafissatou Diouf, F. Drame, Karleen Coly, R. Muhire, Safiatou Thiam, P. Diallo, C. Kane, C. Ndour, E. Volz, Sharmistha Mishra, S. Baral, P. Vickerman (2018)
Estimating the contribution of key populations towards the spread of HIV in Dakar, SenegalJournal of the International AIDS Society, 21
G. Mumtaz, S. Kouyoumjian, N. Hilmi, A. Zidouh, H. Rhilani, K. Alami, A. Bennani, E. Gouws, P. Ghys, L. Abu-Raddad (2013)
The distribution of new HIV infections by mode of exposure in MoroccoSexually Transmitted Infections, 89
U. Marcus, F. Hickson, P. Weatherburn, A. Schmidt (2013)
Estimating the size of the MSM populations for 38 European countries by calculating the survey-surveillance discrepancies (SSD) between self-reported new HIV diagnoses from the European MSM internet survey (EMIS) and surveillance-reported HIV diagnoses among MSM in 2009BMC Public Health, 13
A. Abdul-Quader, A. Baughman, W. Hladik (2014)
Estimating the size of key populations: current status and future possibilitiesCurrent Opinion in HIV and AIDS, 9
B. Weniger, K. Limpakarnjanarat, K. Ungchusak, S. Thanprasertsuk, K. Choopanya, S. Vanichseni, T. Uneklabh, P. Thongcharoen, C. Wasi (1991)
The epidemiology of HIV infection and AIDS in Thailand.AIDS, 5 Suppl 2
A. Okiria, A. Bolo, Victoria Achut, Golda Arkangelo, Acaga Michael, J. Katoro, J. Wesson, S. Gutreuter, Lee Hundley, A. Hakim (2018)
Novel Approaches for Estimating Female Sex Worker Population Size in Conflict-Affected South SudanJMIR Public Health and Surveillance, 5
J. Vandepitte, R. Lyerla, G. Dallabetta, F. Crabbé, M. Alary, A. Buvé (2006)
Estimates of the number of female sex workers in different regions of the worldSexually Transmitted Infections, 82
(2007)
Practical guidelines for intensifying HIV prevention: towards universal access. Geneva: Joint United National Programme on HIV/AIDS
R. Baggaley, B. Owen, R. Silhol, J. Elmes, P. Anton, I. McGowan, A. Straten, B. Shacklett, Que Dang, E. Swann, D. Bolton, M. Boily (2018)
Does per‐act HIV‐1 transmission risk through anal sex vary by gender? An updated systematic review and meta‐analysisAmerican Journal of Reproductive Immunology, 80
P. Wesson, R. Adhikary, A. Jonas, K. Gerndt, A. Mirzazadeh, Frieda Katuta, Andrew Maher, Karen Banda, N. Mutenda, W. Mcfarland, David Lowrance, D. Prybylski, Sadhna Patel (2019)
Estimating the Population Size of Female Sex Workers in Namibia Using a Respondent-Driven Sampling Adjustment to the Reverse Tracking Method: A Novel ApproachJMIR Public Health and Surveillance, 5
Sharmistha Mishra, M. Boily, S. Schwartz, C. Beyrer, J. Blanchard, S. Moses, D. Castor, N. Phaswana-Mafuya, P. Vickerman, F. Drame, M. Alary, S. Baral (2016)
Data and methods to characterize the role of sex work and to inform sex work programs in generalized HIV epidemics: evidence to challenge assumptions.Annals of epidemiology, 26 8
F. Scorgie, M. Chersich, I. Ntaganira, A. Gerbase, F. Lule, Y. Lo (2012)
Socio-Demographic Characteristics and Behavioral Risk Factors of Female Sex Workers in Sub-Saharan Africa: A Systematic ReviewAIDS and Behavior, 16
R. Stuart, Laura Grobicki, H. Haghparast-Bidgoli, J. Panovska-Griffiths, J. Skordis, O. Keiser, J. Estill, Zofia Barańczuk, S. Kelly, Iyanoosh Reporter, D. Kedziora, A. Shattock, J. Petravic, S. Hussain, Kelsey Grantham, R. Gray, Xiao Yap, R. Martin-Hughes, C. Benedikt, N. Fraser-Hurt, Emiko Masaki, David Wilson, M. Gorgens, Elizabeth Mziray, N. Cheikh, Z. Shubber, C. Kerr, D. Wilson (2018)
How should HIV resources be allocated? Lessons learnt from applying Optima HIV in 23 countriesJournal of the International AIDS Society, 21
(2018)
Everyone said no': Biometrics, HIV and Human Rights, a Kenya Case Study
M. Maheu-Giroux, S. Diabaté, M. Boily, N. Jean-Paul, J. Vesga, S. Baral, Kouamé Abo, Venance Wognin, D. Diouf, M. Alary (2019)
Cost-Effectiveness of Accelerated HIV Response Scenarios in Côte d'IvoireJAIDS Journal of Acquired Immune Deficiency Syndromes, 80
(2012)
The potential impact of preexpo- sure prophylaxis for HIV prevention among men who have sex with men and transwomen in Lima, Peru: a mathematical modelling study
E. Fazito, P. Cuchí, M. Mahy, T. Brown (2012)
Analysis of duration of risk behaviour for key populations: a literature reviewSexually Transmitted Infections, 88
M. Ainsworth, W. Teokul (2000)
Breaking the silence: setting realistic priorities for AIDS control in less-developed countriesThe Lancet, 356
K. Sabin, Jinkou Zhao, Jesus Calleja, Yaou Sheng, Sonia Garcia, Annette Reinisch, R. Komatsu (2016)
Availability and Quality of Size Estimations of Female Sex Workers, Men Who Have Sex with Men, People Who Inject Drugs and Transgender Women in Low- and Middle-Income CountriesPLoS ONE, 11
M. Diamond (1993)
Homosexuality and bisexuality in different populationsArchives of Sexual Behavior, 22
P. Vickerman, L. Platt, E. Jolley, T. Rhodes, M. Kazatchkine, A. Latypov (2014)
Controlling HIV among people who inject drugs in Eastern Europe and Central Asia: insights from modeling.The International journal on drug policy, 25 6
T. Brown, W. Peerapatanapokin (2004)
The Asian Epidemic Model: a process model for exploring HIV policy and programme alternatives in AsiaSexually Transmitted Infections, 80
J. Okal, H. Raymond, W. Tun, H. Musyoki, Sufia Dadabhai, D. Broz, J. Nyamu, D. Kuria, N. Muraguri, S. Geibel (2016)
Lessons learned from respondent-driven sampling recruitment in Nairobi: experiences from the fieldBMC Research Notes, 9
G. Mabileau, O. Scutelniciuc, M. Tsereteli, I. Konorazov, A. Yelizaryeva, S. Popovici, K. Saifuddin, E. Losina, M. Manova, V. Saldanha, J. Malkin, Y. Yazdanpanah (2018)
Intervention Packages to Reduce the Impact of HIV and HCV Infections Among People Who Inject Drugs in Eastern Europe and Central Asia: A Modeling and Cost-effectiveness StudyOpen Forum Infectious Diseases, 5
Sharmistha Mishra, M. Pickles, J. Blanchard, S. Moses, M. Boily (2013)
Distinguishing sources of HIV transmission from the distribution of newly acquired HIV infections: why is it important for HIV prevention planning?Sexually Transmitted Infections, 90
J. Williams, M. Alary, C. Lowndes, L. Béhanzin, A. Labbé, S. Anagonou, M. Ndour, I. Minani, C. Ahoussinou, D. Zannou, M. Boily (2014)
Positive Impact of Increases in Condom Use among Female Sex Workers and Clients in a Medium HIV Prevalence Epidemic: Modelling Results from Project SIDA1/2/3 in Cotonou, BeninPLoS ONE, 9
(2000)
Guidelines for second generation HIV Surveillance
(2008)
Commission on
S. Stahlman, T. Sanchez, P. Sullivan, S. Ketende, C. Lyons, M. Charurat, F. Drame, D. Diouf, R. Ezouatchi, S. Kouanda, S. Anato, Tampose Mothopeng, Z. Mnisi, S. Baral (2016)
The Prevalence of Sexual Behavior Stigma Affecting Gay Men and Other Men Who Have Sex with Men Across Sub-Saharan Africa and in the United StatesJMIR Public Health and Surveillance, 2
Peter Piot, A. Seck (2001)
International response to the HIV/AIDS epidemic: planning for success.Bulletin of the World Health Organization, 79 12
David Wilson (2007)
HIV EPIDEMIOLOGY : A REVIEW OF RECENT TRENDS AND LESSONS
Kirsten Stoebenau, L. Heise, J. Wamoyi, N. Bobrova (2016)
Revisiting the understanding of "transactional sex" in sub-Saharan Africa: A review and synthesis of the literature.Social science & medicine, 168
J. Monteiro, S. Galea, T. Flanigan, M. Monteiro, S. Friedman, B. Marshall (2015)
Evaluating HIV prevention strategies for populations in key affected groups: the example of Cabo VerdeInternational Journal of Public Health, 60
A. Bórquez, L. Beletsky, B. Nosyk, S. Strathdee, Alejandro Madrazo, D. Abramovitz, C. Rafful, Mario Morales, J. Cepeda, D. Panagiotoglou, E. Krebs, P. Vickerman, Marie Boily, N. Thomson, N. Martin (2018)
The effect of public health-oriented drug law reform on HIV incidence in people who inject drugs in Tijuana, Mexico: an epidemic modelling study.The Lancet. Public health, 3 9
L. Degenhardt, A. Peacock, Samantha Colledge, J. Leung, J. Grebely, P. Vickerman, J. Stone, E. Cunningham, A. Trickey, Kostyantyn Dumchev, M. Lynskey, P. Griffiths, R. Mattick, M. Hickman, S. Larney (2017)
Global prevalence of injecting drug use and sociodemographic characteristics and prevalence of HIV, HBV, and HCV in people who inject drugs: a multistage systematic reviewThe Lancet. Global Health, 5
A good overview of the primary dynamic models being used for resource allocation and optimization with support from UNAIDS
J. McGillen, S. Anderson, M. Dybul, T. Hallett (2016)
Optimum resource allocation to reduce HIV incidence across sub-Saharan Africa: a mathematical modelling study.The lancet. HIV, 3 9
E. Korenromp, B. Gobet, E. Fazito, J. Lara, L. Bollinger, J. Stover (2015)
Impact and Cost of the HIV/AIDS National Strategic Plan for Mozambique, 2015-2019—Projections with the Spectrum/Goals ModelPLoS ONE, 10
M. Boily, M. Pickles, M. Alary, S. Baral, J. Blanchard, S. Moses, P. Vickerman, Sharmistha Mishra (2015)
What Really Is a Concentrated HIV Epidemic and What Does It Mean for West and Central Africa? Insights From Mathematical ModelingJAIDS Journal of Acquired Immune Deficiency Syndromes, 68
(2008)
Redefining AIDS in Asia, crafting an effective response. Report of the Commission on AIDS in Asia
S. Kelly, S. Kelly, R. Martin-Hughes, R. Stuart, R. Stuart, Xiao Yap, D. Kedziora, D. Kedziora, Kelsey Grantham, S. Hussain, Iyanoosh Reporter, A. Shattock, Laura Grobicki, H. Haghparast-Bidgoli, J. Skordis-Worrall, Zofia Barańczuk, Zofia Barańczuk, Zofia Barańczuk, Olivia Keiser, Olivia Keiser, J. Estill, J. Estill, Janka Petravic, Janka Petravic, R. Gray, C. Benedikt, N. Fraser, M. Gorgens, David Wilson, C. Kerr, C. Kerr, D. Wilson, D. Wilson (2018)
The global Optima HIV allocative efficiency model: targeting resources in efforts to end AIDS.The lancet. HIV, 5 4
J. Colvin, M. Gorgens-Albino, Christine Kasedde (2018)
Analysis of HIV prevention response and modes of HIV transmission : the UNAIDS-GAMET supported synthesis process
K. Lancaster, D. Cernigliaro, R. Zulliger, P. Fleming (2016)
HIV care and treatment experiences among female sex workers living with HIV in sub-Saharan Africa: A systematic reviewAfrican Journal of AIDS Research, 15
Z. Shubber, Sharmistha Mishra, J. Vesga, M. Boily (2014)
The HIV Modes of Transmission model: a systematic review of its findings and adherence to guidelinesJournal of the International AIDS Society, 17
B. Friedland, Laurel Sprague, L. Nyblade, S. Baral, J. Pulerwitz, A. Gottert, U. Amanyeiwe, A. Cheng, C. Mallouris, F. Anam, A. Jackson, S. Geibel (2018)
Measuring intersecting stigma among key populations living with HIV: implementing the people living with HIV Stigma Index 2.0Journal of the International AIDS Society, 21
J. Stover, N. Walker, G. Garnett, J. Salomon, K. Stanecki, P. Ghys, N. Grassly, R. Anderson, B. Schwartländer (2002)
Can we reverse the HIV/AIDS pandemic with an expanded response?The Lancet, 360
S. Goodreau, N. Carnegie, E. Vittinghoff, J. Lama, Jorge Sanchez, B. Grinsztejn, B. Koblin, K. Mayer, S. Buchbinder (2012)
What Drives the US and Peruvian HIV Epidemics in Men Who Have Sex with Men (MSM)?PLoS ONE, 7
(2007)
Practical guidelines for intensifying HIV prevention: towards universal access
(2019)
In: Mendos L. State-Sponsored Homophobia
A. Wirtz, C. Pretorius, C. Beyrer, S. Baral, M. Decker, S. Sherman, M. Sweat, T. Poteat, Jenny Butler, R. Oelrichs, I. Semini, D. Kerrigan (2014)
Epidemic Impacts of a Community Empowerment Intervention for HIV Prevention among Female Sex Workers in Generalized and Concentrated EpidemicsPLoS ONE, 9
S. Larney, A. Peacock, J. Leung, Samantha Colledge, M. Hickman, P. Vickerman, J. Grebely, Kostyantyn Dumchev, P. Griffiths, L. Hines, E. Cunningham, R. Mattick, M. Lynskey, J. Marsden, J. Strang, L. Degenhardt (2017)
Global, regional, and country-level coverage of interventions to prevent and manage HIV and hepatitis C among people who inject drugs: a systematic reviewThe Lancet. Global Health, 5
M. Yotebieng, E. Brazier, Diane Addison, April Kimmel, M. Cornell, O. Keiser, A. Parcesepe, A. Onovo, K. Lancaster, B. Castelnuovo, Pamela Murnane, C. Cohen, R. Vreeman, M. Davies, S. Duda, C. Yiannoutsos, Rose Bono, Robert Agler, C. Bernard, J. Syvertsen, J. Sinayobye, Radhika Wikramanayake, A. Sohn, P. Groote, G. Wandeler, V. Leroy, C. Williams, K. Wools-Kaloustian, D. Nash (2019)
Research priorities to inform “Treat All” policy implementation for people living with HIV in sub‐Saharan Africa: a consensus statement from the International epidemiology Databases to Evaluate AIDS (IeDEA)Journal of the International AIDS Society, 22
M. Boily, R. Baggaley, Lei Wang, B. Mâsse, R. White, R. Hayes, M. Alary (2009)
Heterosexual risk of HIV-1 infection per sexual act: systematic review and meta-analysis of observational studies.The Lancet. Infectious diseases, 9 2
J. Macallister, J. Sherwood, Joshua Galjour, S. Robbins, Jinkou Zhao, K. Dam, A. Grosso, S. Baral (2015)
A Comprehensive Review of Available Epidemiologic and HIV Service Data for Female Sex Workers, Men Who Have Sex With Men, and People Who Inject Drugs in Select West and Central African CountriesJAIDS Journal of Acquired Immune Deficiency Syndromes, 68
J. Burgos, J. Burgos, T. Patterson, J. Graff-Zivin, J. Kahn, M. Rangel, M. Lozada, H. Staines, S. Strathdee (2016)
Cost-Effectiveness of Combined Sexual and Injection Risk Reduction Interventions among Female Sex Workers Who Inject Drugs in Two Very Distinct Mexican Border Cities.PLoS ONE, 11
A good example of how the contribution of sex work to a sub-Saharan African
Sharmistha Mishra, M. Pickles, J. Blanchard, S. Moses, Z. Shubber, M. Boily (2014)
Validation of the Modes of Transmission Model as a Tool to Prioritize HIV Prevention Targets: A Comparative Modelling AnalysisPLoS ONE, 9
A. Low, N. Nagot, I. Konaté, N. Meda, M. Segondy, P. Perre, P. Mayaud, P. Vickerman (2015)
Potential Impact of Existing Interventions and of Antiretroviral Use in Female Sex Workers on Transmission of HIV in Burkina Faso: A Modeling StudyJAIDS Journal of Acquired Immune Deficiency Syndromes, 68
L. Kumaranayake, C. Watts (2001)
Resource allocation and priority setting of HIV|AIDS interventions: addressing the generalized epidemic in sub-Saharan AfricaJournal of International Development, 13
J. Cepeda, K. Eritsyan, P. Vickerman, Alexandra Lyubimova, M. Shegay, V. Odinokova, L. Beletsky, A. Bórquez, M. Hickman, C. Beyrer, N. Martin (2018)
Potential impact of implementing and scaling up harm reduction and antiretroviral therapy on HIV prevalence and mortality and overdose deaths among people who inject drugs in two Russian cities: a modelling study.The lancet. HIV, 5 10
C. Lowndes, M. Alary, Michelyne Belleau, W. Bosu, D. Kintin, Joseph Nnorom, K. Seck, J. Victor-Ahuchogu, D. Wilson (2008)
West Africa HIV/AIDS epidemiology and response synthesis:implications for prevention
P. Luz, B. Osher, B. Grinsztejn, Rachel MacLean, E. Losina, Madeline Stern, C. Struchiner, R. Parker, K. Freedberg, F. Mesquita, R. Walensky, V. Veloso, A. Paltiel (2018)
The cost‐effectiveness of HIV pre‐exposure prophylaxis in men who have sex with men and transgender women at high risk of HIV infection in BrazilJournal of the International AIDS Society, 21
J. Wamoyi, M. Ranganathan, N. Kyegombe, Kirsten Stoebenau (2019)
Improving the Measurement of Transactional Sex in Sub-Saharan Africa: A Critical ReviewJournal of Acquired Immune Deficiency Syndromes (1999), 80
S. Davis, W. Goedel, J. Emerson, B. Guven (2017)
Punitive laws, key population size estimates, and Global AIDS Response Progress Reports: an ecological study of 154 countriesJournal of the International AIDS Society, 20
S. Quaye, H. Raymond, K. Atuahene, R. Amenyah, J. Aberle‐grasse, W. Mcfarland, A. El-Adas, Ghana Group (2015)
Critique and Lessons Learned from using Multiple Methods to Estimate Population Size of Men who have Sex with Men in GhanaAIDS and Behavior, 19
A good example of the improved cost-effectiveness of focusing on key populations in generalized settings when facing cost constraints
A. Rao, S. Schwartz, K. Sabin, T. Wheeler, Jinkou Zhao, J. Hargreaves, S. Baral (2018)
HIV-related data among key populations to inform evidence-based responses: protocol of a systematic reviewSystematic Reviews, 7
An example of an innovative approach to size estimation for MSM using social media to challenge the traditional under-estimation of MSM
(2018)
UNAIDS Data 2018
A. Datta, Wenyi Lin, A. Rao, D. Diouf, A. Kouamé, Jessie Edwards, L. Bao, T. Louis, S. Baral (2019)
Bayesian Estimation of MSM Population Size in Côte d’IvoireStatistics and public policy (Philadelphia, Pa.), 6
K. Shannon, A. Crago, S. Baral, L. Bekker, D. Kerrigan, M. Decker, T. Poteat, A. Wirtz, B. Weir, M. Boily, Jenny Butler, S. Strathdee, C. Beyrer (2018)
The global response and unmet actions for HIV and sex workersThe Lancet, 392
M. Maheu-Giroux, J. Vesga, S. Diabaté, M. Alary, S. Baral, D. Diouf, Kouamé Abo, M. Boily (2017)
Changing Dynamics of HIV Transmission in Côte d'Ivoire: Modeling Who Acquired and Transmitted Infections and Estimating the Impact of Past HIV Interventions (1976–2015)JAIDS Journal of Acquired Immune Deficiency Syndromes, 75
A. Bershteyn, K. Mutai, Adam Akullian, D. Klein, B. Jewell, S. Mwalili (2018)
The influence of mobility among high-risk populations on HIV transmission in Western KenyaInfectious Disease Modelling, 3
E. Volz, N. Ndembi, R. Nowak, G. Kijak, J. Idoko, P. Dakum, W. Royal, S. Baral, M. Dybul, W. Blattner, M. Charurat (2017)
Phylodynamic analysis to inform prevention efforts in mixed HIV epidemicsVirus Evolution, 3
J. Kahn, L. Bollinger, J. Stover, E. Marseille (2017)
Improving the Efficiency of the HIV/AIDS Policy Response: A Guide to Resource Allocation Modeling
S. Kouyoumjian, H. Rhilani, A. Latifi, Amina Kettani, H. Chemaitelly, K. Alami, A. Bennani, L. Abu-Raddad (2017)
Mapping of new HIV infections in Morocco and impact of select interventions.International journal of infectious diseases : IJID : official publication of the International Society for Infectious Diseases, 68
K. Risher, K. Mayer, C. Beyrer (2015)
HIV treatment cascade in MSM, people who inject drugs, and sex workersCurrent Opinion in HIV and AIDS, 10
B. Idrisov, Sean Murphy, Tyler Morrill, Mayada Saadoun, K. Lunze, D. Shepard (2017)
Implementation of methadone therapy for opioid use disorder in Russia – a modeled cost-effectiveness analysisSubstance Abuse Treatment, Prevention, and Policy, 12
Tiffany Lillie, J. Baer, Darrin Adams, Jinkou Zhao, R. Wolf (2018)
Think global, act local: the experience of Global Fund and PEPFAR joint cascade assessments to harmonize and strengthen key population HIV programmes in eight countriesJournal of the International AIDS Society, 21
An example of how MSM in some settings in sub-Saharan Africa may be contributing more to new infections than other populations when downstream transmission is taken into account
Kevin Apodaca, R. Doshi, Moses Ogwal, Herbert Kiyingi, George Aluzimbi, G. Musinguzi, I. Lutalo, Evelyn Akello, W. Hladik (2018)
Capture-Recapture Among Men Who Have Sex With Men and Among Female Sex Workers in 11 Towns in UgandaJMIR Public Health and Surveillance, 5
J. Monteiro, B. Marshall, Daniel Escudero, S. Sosa-Rubí, Andrea González, T. Flanigan, D. Operario, K. Mayer, M. Lurie, O. Galárraga (2015)
Preventing HIV Transmission Among Partners of HIV-Positive Male Sex Workers in Mexico City: A Modeling StudyAIDS and Behavior, 19
(2008)
Analysis of HIV Prevention Re- sponse and Modes of HIV Transmission: The UNAIDS-GAMET Supported Synthesis Process. Johannesburg: UNAIDS Regional Support Team Eastern and Southern Africa
(2017)
:vex014. Demonstrates how phylodynamic modeling work can be used to assess transmission between MSM and female partners
E. Gouws, P. Cuchí (2012)
Focusing the HIV response through estimating the major modes of HIV transmission: a multi-country analysisSexually Transmitted Infections, 88
(2016)
AIDS by the numbers
April Kimmel, Rose Bono, O. Keiser, J. Sinayobye, J. Estill, D. Mujwara, Olga Tymejczyk, D. Nash (2018)
Mathematical modelling to inform ‘treat all’ implementation in sub-Saharan Africa: a scoping reviewJournal of Virus Eradication, 4
(2016)
Fueling the Philippines’ HIV epidemic. Government barriers to condom use by men who have sex with men
Í. Cremin, L. McKinnon, J. Kimani, P. Cherutich, G. Gakii, F. Muriuki, K. Kripke, R. Hecht, Michael Kiragu, Jennifer Smith, W. Hinsley, L. Gelmon, T. Hallett (2017)
PrEP for key populations in combination HIV prevention in Nairobi: a mathematical modelling study.The lancet. HIV, 4 5
G. Gomez, A. Bórquez, C. Caceres, Eddy Segura, R. Grant, G. Garnett, T. Hallett (2012)
The Potential Impact of Pre-Exposure Prophylaxis for HIV Prevention among Men Who Have Sex with Men and Transwomen in Lima, Peru: A Mathematical Modelling StudyPLoS Medicine, 9
(2019)
International Lesbian, Gay, Bisexual, Trans and Intersex AssociationMendos L. State-Sponsored Homophobia 2019
(2011)
The Epidemiology of HIV epidemics in the 21-country West Africa Region: the impact of most at risk populations (MARPs)
(2018)
Miles to go: closing gaps, breaking barriers, righting injustices
A. Bórquez, J. Guanira, P. Revill, Patricia Caballero, A. Silva-Santisteban, S. Kelly, X. Salazar, P. Bracamonte, P. Minaya, T. Hallett, C. Caceres (2019)
The impact and cost-effectiveness of combined HIV prevention scenarios among transgender women sex-workers in Lima, Peru: a mathematical modelling studyThe Lancet. Public health, 4
B. Vuylsteke, L. Sika, Gisèle Semdé, C. Anoma, Elise Kacou, M. Laga (2017)
Estimating the number of female sex workers in Côte d'Ivoire: results and lessons learnedTropical Medicine & International Health, 22
M. Maheu-Giroux, J. Vesga, S. Diabaté, M. Alary, S. Baral, D. Diouf, Kouamé Abo, M. Boily (2017)
Population-level impact of an accelerated HIV response plan to reach the UNAIDS 90-90-90 target in Côte d’Ivoire: Insights from mathematical modelingPLoS Medicine, 14
Introduces a critically important effort to collate and systematize key population data globally to make it more available for modeling, analysis and decision-making
(2019)
Cost-effectiveness of acceler- ated HIV response scenarios in Cote d’Ivoire
(2010)
New HIV Infections by mode of transmission in West Africa: A Multi-Country Analysis. Dakar, Senegal: UNAIDS Regional Support Team for West and Central Africa
(2016)
Optimum resource alloca- tion to reduce HIV incidence across sub-Saharan Africa: a mathematical modelling study. Lancet HIV 2016; 3:e441–e448
M. Kavanagh, S. Baral, M. Milanga, J. Sugarman (2019)
Biometrics and public health surveillance in criminalised and key populations: policy, ethics, and human rights considerations.The lancet. HIV
E. Broughton, Ó. Núñez, R. Arana, Alexey Oviedo (2016)
Effectiveness and Efficiency of Improving HIV Service Provision for Key Populations in NicaraguaFrontiers in Public Health, 4
(2016)
Fueling the Philippines' HIV epidemic. Government barriers to condom use by men who have sex with men. USA: Human Rights Watch
Downloaded from https://journals.lww.com/co-hivandaids by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD35y8U/jUqeEyVMweXB0LtB1dl03evjuXlpE8l2qkMB64= on 12/16/2019 REVIEW URRENT Evolving HIV epidemics: the urgent need to refocus PINION on populations with risk Tim Brown and Wiwat Peerapatanapokin Purpose of review To explore the comparative importance of HIV infections among key populations and their intimate partners as HIV epidemics evolve, and to review implications for guiding responses. Recent findings Even as concentrated epidemics evolve, new infections among current and former key population members and their intimate partners dominate new infections. Prevalent infections in the general population grow primarily because of key population turnover and infections among their intimate partners. In generalized epidemic settings, data and analysis on key populations are often inadequate to assess the impact of key population-focused responses, so they remain limited in coverage and under resourced. Models must incorporate downstream infections in comparing impacts of alternative responses. Summary Recognize that every epidemic is unique, moving beyond the overly simplistic concentrated/generalized epidemic paradigm that can misdirect resources. Guide HIV responses by gathering and using locally relevant data, understanding risk heterogeneity, and applying modeling at both national and sub-national levels to optimize resource allocations among different populations for greatest impact. Translate this improved understanding into clear, unequivocal advice for policymakers on where to focus for impact, breaking them free of the generalized/concentrated paradigm limiting their thinking and affecting their decisions. Keywords focused intervention, HIV modeling, HIV prevention, key populations, national responses INTRODUCTION: CHANGING RESPONSES pregnant women) and generalized epidemics (con- TO EVOLVING EPIDEMICS sistently over 1% in pregnant women). As the concentrated/generalized paradigm was In the late 1990s, UNAIDS and WHO introduced the simple, easily defined by prevalence and heavily concept of low-level, concentrated and generalized promoted by international technical partners, it epidemics as a tool to guide surveillance strategies was quickly adopted on a large scale. By the mid- [1]. Although the concept of low-level epidemics fell 2000s, it was also being used to define national by the wayside as HIV became globally ubiquitous, responses. Countries with concentrated epidemics the paradigm of concentrated and generalized epi- were told to focus on locally relevant at-risk pop- demics continues to dominate many people’s think- ulations; countries with generalized epidemics were ing about HIV epidemics and responses to them. The to put most of the emphasis on the general popula- original definition distinguished these two epi- tion [2]. In the early 2000s, the challenge in taking demic types by whether there was sustained HIV transmission within the ‘general population’, that is, that part of the population outside of certain East-West Center, Honolulu, Hawaii, USA groups perceived to be at high risk of acquiring Correspondence to Tim Brown, Research Program, East-West Center, HIV. These are groups, which are normally referred 1601 East-West Road, Honolulu, HI 96848, USA. Tel: +1 808 944 7476; to today as ‘key populations’, and include: MSM, e-mail: tim@hawaii.edu male and female sex workers (MSW and FSW) and Curr Opin HIV AIDS 2019, 14:337–353 their clients, people who inject drugs (PWID), trans- DOI:10.1097/COH.0000000000000571 gendered individuals and prisoners and other incar- This is an open access article distributed under the Creative Commons cerated individuals. UNAIDS and WHO also defined Attribution License 4.0 (CCBY), which permits unrestricted use, distri- simple numerical proxies for concentrated (>5% in bution, and reproduction in any medium, provided the original work is at least one subpopulation, but <1% among urban properly cited. 1746-630X Copyright 2019 The Author(s). Published by Wolters Kluwer Health, Inc. www.co-hivandaids.com Downloaded from https://journals.lww.com/co-hivandaids by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD35y8U/jUqeEyVMweXB0LtB1dl03evjuXlpE8l2qkMB64= on 12/16/2019 Concentrated epidemics people blame these policy failures on the govern- KEY POINTS ment’s focus on HIV prevention policies that target heterosexual couples rather than members of the Most concentrated epidemics do not go generalized, LGBT community’ [7]. This interpretation of epi- and key populations do matter in demic dynamics is further buttressed by the obser- generalized epidemics. vation that countries with long running Turnover from key populations and transmission to their concentrated epidemics have an increasing num- intimate partners drive the increasing proportion of ber of HIV cases detected or presenting for treat- prevalent infections outside key populations as ment from the general population. Coupled with concentrated epidemics age. widespread stigma and discrimination, this belief Assessing the contribution of program responses among in a transitioning epidemic encourages redirection key populations to averting new infections and of resources and makes it challenging to obtain preventing HIV-related deaths requires using locally national/local resources for robust responses calibrated, dynamical models that capture among key populations. downstream effects. Countries with generalized epidemics, on the The influence of key populations on the epidemics of other hand, were encouraged to focus most of their sub-Saharan Africa has been underestimated, leading resources on the general population. This led to to inefficient resource allocations and less programs in sub-Saharan Africa that focused almost effective programs. entirely on the population at large with some lim- The generalized/concentrated paradigm should be ited attention to heterosexual transmission through retired in favor of recognizing that every epidemic is sex work [8]. Although the 2007 UNAIDS guidelines unique, using data and model-based understanding of on intensifying prevention in generalized settings local risk variations to better guide programs. explicitly state that ‘programmes for most-at-risk populations remain important’, they focus primar- ily on general population efforts. Consequently, this advice in concentrated settings was that UNAIDS given people’s challenges in absorbing complex and other partners were stressing the need for pre- messages, resources flowed to general population vention and treatment to reach everyone and pro- efforts whereas key population programs in gener- moting an expanded response of prevention, alized epidemics remained limited in scope, funding treatment and mitigation efforts covering the entire and coverage. This persists today [9–12]. The corol- population [3,4]. This was an essential part of build- lary to ‘concentrated becomes generalized’ is ‘once ing the successful global coalition to expand and epidemics become generalized, key populations mobilize funding for the global response in the aren’t too important’. This has become a self-fulfill- decade after the 2000 Durban AIDS Conference, ing prophecy as the lack of interest in key popula- but it often resulted in failures to prioritize programs, tions in generalized settings has restricted the especially programs for key populations [5]. Given collection of data on prevalence, size and risk behav- the challenges in mobilizing resources for socially iors until recently, making it difficult to accurately and politically stigmatized populations in this envi- assess the proportion of new infections occurring ronment, advocates frequently seized upon a ‘con- among key populations and their immediate part- && centrated epidemics will become generalized’ ners [13 ]. argument to build support for key population efforts. Today, this underlying misconception of ‘con- Goal of this review centrated epidemics will become generalized’ leads the press, some advocates and many decision The goal of this review is to improve understanding makers controlling national resources to frame con- of how to focus responses in different epidemic centrated epidemics in terms of transitioning ‘het- settings for maximum effectiveness. It, thus, erosexual’ or ‘general population’ epidemics and to addresses three directly relevant questions arising make program choices accordingly. For example: ‘In from the discussion above: Eastern Europe, heterosexual transmission now accounts for 55% of new infections. So the epidemic (1) Given increasing general population case is moving from key populations to the general pop- reports in long-running concentrated epidem- ulation’ [6]. Even when followed by a call for key ics, should the response emphasis shift from key population programs, such statements reinforce a populations to the general population? prevailing paradigm that can lead to bad program (2) How can the influence of key populations in choices, for example, ‘Philippine rights activists for generalized settings be characterized and are lesbian, gay, bisexual and transgender (LGBT) current responses adequate? 338 www.co-hivandaids.com Volume 14 Number 5 September 2019 Evolving HIV epidemics Brown and Peerapatanapokin (3) What do data, models and analyses say about most impactful combinations. The second impor- these two questions and how can they be tant observation is that countries with the longest applied to better inform key decision makers running epidemics, that is, Thailand and Cambodia, of the role of key populations in local have the largest proportion of current infections epidemics? among the general population, shown on the graphs as ‘Rest of males/females’. They also show a large portion of new infections, roughly 50%, among the general population. In the more recent and growing THE EVOLUTION OF CONCENTRATED epidemics, for example those in Pakistan and the EPIDEMICS: A REGIONAL CASE STUDY Philippines, most new infections occur among key Why does the proportion of reported infections populations and a smaller proportion of current from the general population increase in long-run- infections are found among the general population. ning concentrated epidemics and what are its impli- Figure 2 shows the combined growth of current cations for prevention efforts? Models are best and new infections for these 11 countries. At a positioned to answer this question. The AIDS Epi- regional level, the proportion of current infections demic Model (AEM) is a concentrated epidemic among the general population grows steadily as model originally developed for Asian countries epidemics run longer; by 2018, 65% of current [14]. AEM contains most key populations: male infections are outside key populations. However, and female sex workers and their clients, MSM, the proportion of new infections outside key pop- people who inject drugs and transgendered popu- ulations never rises above 35%. The other important lations. It also includes nonkey population men and observation about HIV in the general population is women. AEMmodelsthe various routes oftrans- that while current infections are almost equally mission among and between these populations, divided between men and women, new infections including heterosexual transmission through non- among women outnumber new infections among commercial casual sex and sex between intimate men by a factor of 3 to 1. This raises several impor- partners. AEM is widely used throughout Asia; 11 tant questions: countries used it to prepare the national models submitted to UNAIDS in 2018. AEM tracks current (1) With two-thirds of new HIV infections occur- and new infections, sources of infection, and turn- ring among key populations, how can two- over for each population in the model, providing a thirds of current infections be found outside regional ‘laboratory’ for exploring the natural his- them? tory of these epidemics. (2) Why does the proportion of current infections outside key populations increase with epidemic age? Is transmission in these epidemics ‘going The evolving epidemics of Asia generalized’ as many believe? Figure 1 shows the number of (a) current infections (3) Why is the number of new infections among and (b) new infections in the 11 countries in 2017 by women outside key populations three times that population. The proportion of current infections for men? outside key populations varies from 18 to 86%, whereas the proportion of new infections arising outside key populations varies from 10 to 62%. How key populations influence epidemic The first thing of note is the incredible diversity dynamics in patterns of new infections by key population between the countries. In five countries, people Answering these questions requires a deeper insight who inject drugs account for one-third or more of into the varied roles key populations play in the new infections. In the Philippines, the epidemic is dynamics of concentrated epidemics. Key popula- almost completely among MSM, whereas in other tions influence the evolution of these epidemics in countries, the proportion of new infections among four ways: MSM is smaller but growing. New infections among First, if no protective steps are taken, HIV trans- sex workers and clients now account for only 19% of mission among key population members is highly new infections in these countries, although it varies efficient. Transmission risk in key populations is from 0.1 to 35% across the countries shown. Pre- elevated for several reasons: the high-efficiency of vention resources must be allocated differently in anal sex in transmitting HIV for MSM and transgen- these countries to maximize their impact and mul- dered populations [15]; high-frequency of needle tiple AEM scenarios for different program mixes are sharing acts among many PWID; substantial partner typically prepared to inform decision makers of the exchange rates and the facilitating role of other 1746-630X Copyright 2019 The Author(s). Published by Wolters Kluwer Health, Inc. www.co-hivandaids.com 339 Concentrated epidemics (a) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% (b) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% FIGURE 1. (a) The number of prevalent, that is, current, HIV infections in 11 Asian countries in 2017 by subpopulation, and (b) the number of new HIV infections by subpopulation. The countries have been ordered from left to right based on the proportion of current infections among key populations. The figures are derived from national models developed by local AEM teams as submitted to UNAIDS in 2018. Values are for female sex workers (FSW), clients, MSM, transgenders (TG), people who inject drugs (PWID) and the rest of the male and female population. AEM, AIDS Epidemic Model. sexually transmitted infections and primary HIV returns home to marry and open a shop. Young infection for sex workers and clients. Vaginal sex, men may be particularly active as clients, but then especially with much lower rates of facilitating sex- stop visiting sex workers after marriage. Many PWID ually transmitted infections outside certain key pop- stop injecting after several years. At this point, these ulations, transmits HIV at significantly lower rates former members of key populations return to the and over a much longer time span [16]. These factors ‘general population’. Infections that occurred while give key populations a disproportionate role in epi- they were at elevated risk may be detected long after demic dynamics, thus their larger contribution to they have re-entered that general population. AEM new infections. tracks this turnover between populations. Figure 3a Second, key populations are not closed; people shows the proportion of HIV infections occurring move in and out of them [17]. A young woman may within key populations that have since returned to become a sex worker for several years, but then the general population for Thailand, one of the 340 www.co-hivandaids.com Volume 14 Number 5 September 2019 Evolving HIV epidemics Brown and Peerapatanapokin (a) (b) 2,000,000 200,000 1,800,000 180,000 1,600,000 160,000 1,400,000 140,000 1,200,000 120,000 1,000,000 100,000 800,000 80,000 600,000 60,000 400,000 40,000 200,000 20,000 0 0 FIGURE 2. The evolution of (a) current infections and (b) new infections by subpopulation from 1990 to 2020 aggregated across the 11 AIDS Epidemic Model countries. The gray bars show the evolution of current and new infections in the general population (rest of males/rest of females). longest-running epidemics in Asia. Through 2017 increase in prevalent infections in the general pop- almost 502 000 HIV-positive men who contracted ulation: vastly more ex-members of key populations HIV while in key populations have returned to the are men and pass HIV onto their spouses, contrib- general population, whereas only 52 000 new infec- uting to an increase in prevalent infections among tions have occurred directly within the male general general population women. population. Cumulatively, 91% of the infections Fourth, the infection of wives and other inti- found among the male general population origi- mate partners is an example of how infections nated from transmission within key populations. occurring among key populations can result in sub- For women, the effect is less pronounced as cumu- stantial numbers of downstream transmissions. latively only 71 000 HIV-positive sex workers Figure 4 shows the impact of averting 1000 infec- returned to the female general population compared tions among female sex workers in 2018 for with 311 000 infections occurring among these Indonesia. For the first few years, most of the averted women, that is, 19%. This movement of HIV-posi- infections are among clients. However, as clients tive individuals from key populations partially and ex-clients then more gradually transmit HIV explains the growing proportion of general popula- to their current and future wives, the number of tion prevalent infections as epidemic ages. In terms infections averted among women not in key pop- of newly detected infections in each population, the ulations increases over time, a pattern first observed effect appears even larger as many older infections by Weniger et al. [18] in 1991 in Thailand that has are only diagnosed after progression to symptomatic repeated throughout Asia. When there is a substan- illness, that is, long after the person leaves a tial nexus between injecting drug use and sex work, key population. an epidemic among PWID can jump start the sex Third, members of key populations often have work component of an epidemic, producing large intimate partners to whom HIV may be transmitted, numbers of downstream infections [19,20]. both while in key populations and after transition- These dynamic factors provide answers to the ing out of them. For the 11 AEM countries, Fig. 3b questions raised earlier. Most of the new infections shows the largest proportion of new infections occur in key populations because of the more effi- among women not in key populations comes from cient transmission among them, but most of those sex with their current spouse; however, as described infections eventually become prevalent general in the preceding paragraph, most infections among population infections through turnover. In addi- these men were acquired when they were members tion, HIV is transmitted to their predominantly of key populations. This is the other piece of the female intimate partners who are not themselves 1746-630X Copyright 2019 The Author(s). Published by Wolters Kluwer Health, Inc. www.co-hivandaids.com 341 Concentrated epidemics (b) 50,000 40,000 30,000 (a) 500,000 20,000 400,000 300,000 10,000 200,000 100,000 2000 2005 2010 2015 2020 Clients FSW PWID MSM Sex with PWID Sex with MSM Total infecons Le KP Casual sex Sex with spouse FIGURE 3. (a) The number of HIVþ individuals leaving each key population to return to the ‘general population’ in Thailand from the start of the epidemic to 2017, and (b) infections among nonkey population women in the 11 countries over time by route of infection. key population members. In the countries modeled Increasing rates of ART coverage, which greatly here only 2.5% of cumulative HIV infections extend people’s lives, increase this effect. Thus, through 2017 occurred through casual sex, that is, the longer an epidemic runs the greater the propor- sex outside of relationships that were not sex work- tion of current infections that are outside key pop- related. The proportion of current infections outside ulations. New infections among women outside key key populations increases over time because turn- populations greatly outnumber those of men over accumulates faster than mortality effects. because most arise from male partners who are or Cumulave downstream infecons averted: remove 1000 FSW infecons in Indonesia in 2018 18,000 16,000 14,000 12,000 10,000 8,000 6,000 4,000 2,000 FIGURE 4. Downstream infections averted in the Indonesia model if 1000 female sex worker infections are prevented in 2018. By 2030, this will avert 9200 infections among clients, 640 additional infections among sex workers, 420 infections among males not in key populations and 4200 infections among females not in key populations. 342 www.co-hivandaids.com Volume 14 Number 5 September 2019 Infecons among non-KP women Evolving HIV epidemics Brown and Peerapatanapokin were members of key populations and the number current distributions of new infections, but also of men in key populations vastly exceeds the num- consider downstream infections in identifying pro- ber of women. In the 11 countries described here, grams with the greatest impact. there were 24 million clients, 740 000 PWID, 4 million MSM, but only 1 million FSW. CONCENTRATED EPIDEMICS IN OTHER This picture has important implications for pre- REGIONS: SIMILAR BUT DIFFERENT vention efforts in today’s concentrated settings. First, because most infections in the general popu- Globally, the ‘concentrated’ regions are Asia and lation originate from infections within key popula- thePacific,the MiddleEastand NorthAfrica tions or transmission to their intimate partners, a (MENA), Eastern Europe and Central Asia (EECA), shift to general population prevention programs and Western and Central Europe and North Amer- would be a major mistake. Instead, the focus must ica (WCENA). Latin America and the Caribbean be kept on key population prevention; this will have (LAC) are a mix with Latin America being predomi- the greatest downstream impact on total infections, nantly concentrated epidemics, whereas the Carib- including those in the general population, espe- bean has more generalized epidemics. Does the cially women. Second, programs for intimate part- situation described for Asia apply in other concen- ners of key populations should be an essential trated epidemic settings, that is, are the dynamics component of all prevention efforts. It would also likely to be the same in the other regions considered be valuable to develop innovative programs focused concentrated? on encouraging former key population members to UNAIDS analyzed the data from national mod- test and, if positive, begin antiretroviral treatment els to estimate the proportion of new infections in and be supported in offering testing to partners 2017 occurring among key populations and their through assisted partner notification services. These immediate partners in each region, the result is efforts would address most infections occurring out- shown in Fig. 5 [21]. The combined proportions side key populations. Finally, in allocating resources for sex workers, clients and key population partners for prevention programs, look not only at the are quite similar for Asia and the Pacific, EECA and EASTERN EUROPE & WESTERN & CENTRAL ASIA & PACIFIC AFRICA CENTRAL ASIA 4% 3% 2% 9% 16% 16% 35% 39% 14% 28% 12% 60% 2% 10% 29% 21% MIDDLE EAST & EASTERN & SOUTHERN LATIN AMERICA NORTH AFRICA AFRICA 3% 2% 2% 8% 13% 23% 6% 24% 1% 38% 3% 30% 6% 83% 41% 17% SW Clients and KP partners MSM TG PWID Rest of populaon FIGURE 5. Distribution of new infections by sub-population in different regions of the world in 2017 as estimated by UNAIDS. Source: Miles to Go 2018 [21]. 1746-630X Copyright 2019 The Author(s). Published by Wolters Kluwer Health, Inc. www.co-hivandaids.com 343 Concentrated epidemics MENA (37–43%). Latin America has a smaller pro- populations themselves are more poorly defined portion from these groups, but almost twice the and less visible than in concentrated settings. For proportion from MSM, whereas the epidemics in example, ‘sex work’ often involves a mix of self- EECA and MENA are much more strongly influ- identified, more active, professional sex workers in enced by PWID than the other regions. Latin Amer- diverse settings [41] and less easily identifiable, ica also shows a significant proportion among more occasional, women engaging in transactional transgendered populations. In WCENA (not sex [42,43]. MSM in SSA face severe stigma and shown), the proportion among MSM is even larger, discrimination, keeping them out of sight and lim- 57%. iting programs, and the situation is worsening with Work with dynamical models in these regions the regional resurgence of conservative attitudes has tended to focus on the most important key [44,45]. Although there is growing awareness, with populations in the region or country as revealed 36 out of 47 countries reporting some evidence of by surveillance data. The work in Latin America injecting drug use in 2017 [46], only 7 countries had has primarily explored programs for MSM [22–27] any needle and syringe programs for people who and transgender women [28], although some inject drugs [9]. modeling work has looked at FSW and PWID [29– 31]. In EECA, the modeling is strongly skewed Comparative prevalence among key toward PWID populations, given the dominant role populations is high whereas size estimates of injecting in new infections [32–35]. And in are often low MENA, modeling work has focused on injecting [36] with some modes of transmission work in In recent years, key population data collection has Morocco, which did highlight significant new infec- increased, revealing that members of key popula- tions from sex work [37,38]. tions in the region are at elevated risk for HIV. Most of these efforts did not use models incor- Figure 6 compares HIV prevalence for FSW and porating all populations but focused on analysis of MSM as reported to UNAIDS with national preva- effectiveness and cost-effectiveness of programs lence [47]. The Eastern and Southern Africa (ESA) within specific populations, such as MSM or PWID. region reports some of the highest prevalence However, there has been substantial work done with among sex workers in the world, whereas reports && Optima in EECA and Latin America [39 ]. Optima is in Western and Central Africa (WCA) are lower, but a full dynamical model, incorporating the effects of still substantial. The median ratio of FSW to national downstream impacts in identifying the most effec- prevalence in ESA and WCA is 5 and 7, respectively. tive programs [40]. In normal country applications, Prevalence among MSM is high in both regions, but both key populations and the general population with a higher median ratio for MSM to national are included. If HIV spread outside key populations prevalence of 8 in WCA versus 2 in ESA. Among were becoming a significant contributor, one would PWID, prevalence in the 4 ESA countries reporting expect these analyses to recommend increases in data varies from 8.5 to 46.4%. Although 10 WCA prevention efforts, for example, condom promotion countries report generally lower prevalence among among the general population. This was not seen – PWID, varying from 1.6 to 10.2% [47]. in fact, the recommendations in 11 countries in Another critical input in assessing the contribu- EECA and 4 countries in Latin America were to scale tion of key populations in SSA is their size, shown as up or sustain one or more key population programs a proportion of the population in Fig. 7. Standard while scaling down general population and youth techniques for estimating key population size, for && programs [39 ]. Given these results and the general example, census/mapping, multiplier or capture– similarity of the regional patterns of new infections recapture [48], present their own challenges in in epidemics that have been ongoing for two to sub-Saharan Africa. Resource constraints on key three decades, it seems unlikely that transmission population programs often limit the ability to do outside key populations and their intimate partners frequent or large-scale mapping. Multiplier methods plays a major role. can underestimate numbers if key population pro- gram data are limited or populations are unrecog- nized and/or inaccessible to surveyors, for example, THE CHALLENGES OF ESTIMATING KEY when stigmatized population members misreport POPULATION EFFECTS IN GENERALIZED their risk. These issues have been observed in size SETTINGS estimation estimates in the region for both MSM Evaluating the role of key populations in the and sex workers [49,50]. Serious efforts to improve generalized epidemic settings of sub-Saharan Africa size estimates have often used multiple methods (SSA) is more challenging. In some cases, the key [51–53] or applied innovative approaches, for 344 www.co-hivandaids.com Volume 14 Number 5 September 2019 Evolving HIV epidemics Brown and Peerapatanapokin (a) Prevalence among FSW compared to naonal 15-49 prevalence 80.0% 72% 70.0% ESA WCA 60% 56% 60.0% 50.0% 46% 40.0% 28% 30.0% 24% 24% 17% 20.0% 13% 10% 11% 9% 7% 6% 10.0% 5% 4% 0.0% Naonal FSW (ESA) (b) Prevalence among MSM compared to naonal 15-49 prevalence 50.0% 44% 45.0% 42% ESA WCA 40.0% 33% 35.0% 31% 30.0% 27% 23% 25.0% 22% 18% 18% 20.0% 17% 15% 14% 14% 13% 12% 15.0% 11% 10.0% 6% 4% 3% 5.0% 2% 0.0% Naonal MSM(ESA) MSM(WCA) FIGURE 6. HIV prevalence among (a) Female sex worker and (b) MSM compared against national prevalence among those aged 15–49 years for Eastern and Southern Africa (ESA in orange) and Western and Central Africa (WCA in blue). Although not nationally representative in most cases, they give an idea of the range of values being observed in the regions. Source: UNAIDS Data 2018 [47]. example, Bayesian techniques applied to RDS multi- information from program clients on their key pop- pliers [54] or combining venue-based sampling and ulation status, storing this information so as to RDS [55]. ensure confidentiality and prevent abuse, and devel- The increasing global emphasis on quality pro- oping unique identifiers for service improvement gram data [56] presents another potential source for and de-duplication [57]. High levels of discrimina- size estimates, but it requires de-duplication across tion and social stigma almost inevitably lead to multiple service providers and venues. However, under-reporting and low size estimates [58]. this faces major hurdles: obtaining accurate Although biometrics have great potential as unique 1746-630X Copyright 2019 The Author(s). Published by Wolters Kluwer Health, Inc. www.co-hivandaids.com 345 Concentrated epidemics FSW esmates as proporon of 15-49 women (a) 3.0% 2.68% WCA 2.5% ESA 2.05% 2.0% 1.5% 1.28% 1.17% 1.07% 0.88% 1.0% 0.75% 0.68% 0.62% 0.49% 0.48% 0.36% 0.5% 0.23% 0.10% 0.07% 0.03% 0.0% MSM esmates as proporon of 15-49 men (b) 2.5% 2.17% 1.94% WCA ESA 2.0% 1.59% 1.5% 1.01% 1.0% 0.74% 0.46% 0.42% 0.43% 0.5% 0.29% 0.24% 0.22% 0.15% 0.08% 0.09% 0.03% 0.03% 0.02% 0.0% FIGURE 7. Size estimates for (a) Sex worker and (b) MSM as a proportion of 15–49-year-old population of the same sex for Eastern and Southern Africa (ESA in orange) and Western and Central Africa (WCA in blue). Source: UNAIDS Data 2018 [47]. identifiers, they have sometimes been rejected by ranges [60] and data from more recent reviews affected communities because they fear abuse by [61]. However, in both regions, estimates for MSM authorities [59]. tend to be quite low, mostly less than 0.5% of 15–49 Despite these challenges, the number of size men, and in several cases less than 0.1%. These are estimates is growing. On average, ESA countries much lower than are biologically plausible [62] or providing estimates report larger proportions of are seen in more systematic data collection in other the 15–49 population being FSW or MSM than regions [63]. Social media-based estimates of same- WCA countries. The estimates for FSW are mostly sex interest in Africa also give much higher esti- in agreement with past reports of regional size mates than reported to UNAIDS [64 ]. As Davis et al. 346 www.co-hivandaids.com Volume 14 Number 5 September 2019 Evolving HIV epidemics Brown and Peerapatanapokin [58] have highlighted, countries criminalizing epidemics as ones where transmission would not be same-sex behavior often report lower proportions sustained in the absence of key populations, such as of MSM in the adult population, and many of the sex workers, clients, MSM and PWID [71] and to the countries of sub-Saharan Africa criminalize and/or increasing use of the term ‘mixed’ epidemics by exhibit extensive homophobia [65]. Estimates of others, where both key population and general pop- people who inject drugs for the region vary from ulation transmission were active [72]. In either case, 0.14 to 1.0% of males aged 15–64 years, with 11.6% many began thinking of the epidemics in some of PWID reported to be women [46]. WCA countries as concentrated and/or mixed, but others without in-depth understanding of the epi- demiology and what the data were telling them Early attempts underestimated the influence continued to think of them as generalized. of key populations in generalized settings In the mid-2010s, the MOT model was correctly The earliest widespread estimates of the impact of criticized as a ‘static’ model, giving only a cross- key populations in SSA occurred in the mid-2000s sectional snapshot of incidence and failing to cap- using the Modes of Transmission (MOT) model, ture the true population level impact of interven- which estimates the proportion of annual new infec- tions to reduce new infections among key tions acquired by sub-populations based on preva- populations [73–75]. Another critique was of its lence, size and some limited behavioral data [66]. focus on those acquiring HIV rather than those MOT studies in ESA showed between 60 and 95% of transmitting it, which were the logical place to new infections occurring among the ‘general het- direct interventions. Thus, it had limitations in erosexual population’ with another 7–15% attrib- allocating prevention and treatment resources uted to sex work. Contributions from PWID and among different sub-populations to maximally MSM were at most a few percent. In Western Africa, reduce future infections and deaths. The core of this a higher contribution of new infections because of criticism was that the MOT model did not include sex work was observed, 10–32%, but still 54–72% of the downstream benefits in future averted infections new infections were among the general population. that accrue from preventing an infection in a key Contributions from key populations were larger population today, as discussed earlier. than in ESA, with between 1 and 12% among MSM and 1–8% among PWID. These estimates Dynamic models of the influence of key depend critically on both prevalence and size esti- populations change the picture mates, meaning the limited availability of represen- tative prevalence data and the skew toward lower Mishra et al. [76] reviewed the importance of incor- size estimates for some populations may misrepre- porating downstream infections in assessing the sent the magnitude of contributions. It is worth impacts of sex work interventions in generalized noting that the MOT results showed substantial epidemic settings. They found median estimates of inter-country variation in the contributions of dif- the contribution of sex work to the epidemic, com- ferent groups to new infections, just as AEM did bining infections among sex workers and clients in Asia. from static, MOT-style exercises (9%) were much From the late 2000s through the mid-2010s, lower than those coming from a limited set of thinking about the epidemics in West and Central dynamical models available at the time (14–38% Africa evolved. Although ‘generalized’ by the 1% in with sex work interventions and 58–89% without) pregnant women definition with MOT analyses [77–79]. Accordingly, they recommended sum- showing the most new infections among general ming new infections in sex workers and clients with population categories [67,68], closer examination of the resulting downstream infections from chains of local data in many countries of the region called transmission to others over many years to more into question whether the heterosexual compo- fully capture the longer term benefits of preventing nents of the epidemic were self-sustaining in the sex work infections today. This is in fact what is absence of key populations [69]. Circumcision was routinely done now in cost-effectiveness and more common in WCA than in ESA, which tends to resource allocation exercises in sub-Saharan Africa. reduce transmission, and national prevalence was Dynamical models, such as Goals, Optima and generally lower. It was also recognized that the EMOD are used to estimate the total number of traditional generalized/concentrated prevalence infections and/or deaths in all populations over distinction failed to align programming with epide- some extended time frame, for example, 10, 20 or miology in the countries of WCA, which called for 30 years, under alternative resource allocation sce- more, but not exclusive, emphasis on key popula- narios, including those addressing key populations && && tions [70]. This led some to redefine concentrated [39 ,80 ]. Comparing such impact estimates 1746-630X Copyright 2019 The Author(s). Published by Wolters Kluwer Health, Inc. www.co-hivandaids.com 347 Concentrated epidemics inherently incorporates downstream infections prevention efforts for and high ART coverage into identifying the most efficient and effective among FSW. Cremin et al. [83] modeled Nairobi, intervention mix. However, doing this comes at Kenya where the overall epidemic is in decline, but thecostofmeeting themorecomplex input data an epidemic persists among MSM and male sex requirements of dynamical models, including prev- workers (MSW). Their in-depth model, based on alence, behaviors and size estimates, and calibrat- extensive data collation in Nairobi and fitting to ing the model for trends over time in the historical data in each group, included various het- various populations. erosexual groups, multiple risk categories for sex Retrospective modeling including downstream workers, and MSM, but excluded PWID. Given con- infections makes the important early contributions tinuing declines in incidence among heterosexuals of sex work to generalized epidemics apparent. In and sex workers, their analysis identified the opti- modeling Kisumu, Kenya from 2000 to 2020 with mal portfolio for prevention as one focused first on STDSIM, Steen et al. [78] found that removing sex condom promotion for MSM and MSW followed by work entirely would have reduced incidence by 66% strengthening ART efforts and then other interven- and prevalence by 56%, but still resulted in a self- tions. Maheu-Giroux et al. [81 ] did a similar exer- sustaining heterosexual epidemic. Sex workers were cise in Cote d’Ivoire constructing a detailed divided into high, medium and low contacts, with historical model, also excluding PWID, that fit the high group being more visible sex workers and observed trends. They found that achieving 90– the low contact group representing women engag- 90–90 targets among sex workers, clients and ing in transactional sex. Of interest, they found that MSM alone could achieve a 30% reduction in intervention among the high group alone accrued new infections, compared with 50% if the targets almost the same benefits as reaching all sex workers. are reached for all [84]. This key population-focused A similar analysis in Coˆte d’Ivoire showed that over approach was the most cost-effective of multiple && time, the 10-year population attributable fraction scenarios examined [85 ]. Recently phylodynamic (PAF) of infections because of sex work dropped approaches are being combined with modeling to from 95% in 1976–1985 to about 19% in 2005– look at the downstream impact of key populations. && 2015, decreasing as the epidemic became more Volz et al. [86 ] explored the influence of MSM on established in the heterosexual population [81 ]. the epidemic in Abuja, Nigeria, finding that 9% of Another model for Bobo-Dioulasso, Burkina Faso infections of women were from partners who were found the 5-year PAF for sex work dropping from MSM in 2014. Because of downstream effects, they 75 to 88% from 1985 to 1990 to 39% from 1995 to found that focused treatment for MSM could avert 2010 as condom use rose among sex workers [77]. 27% of infections over 20 years, compared with The study also found a higher PAF of 60–70% asso- 54% for universal test and treat, making a targeted ciated with full-time sex workers compared with 10– approach much more cost-effective. 20% for occasional sex workers, reinforcing the Outside of Goals and Optima work, little has findings of Steen et al. been done on dynamical models for people who Other than Goals-based and Optima-based inject drugs, mirroring the limited PWID programs models in the region, which routinely include all noted by Larney et al. [9]. Rhodes et al. [87] built a key populations whenever data are available, other model to look at the effects of methadone programs dynamical models including MSM and PWID are in Kenya, but it had a limited sexual transmission rarer than those that focus on sex work and much model that did not situate PWID in terms of overall more geographically limited. However, as inci- influence on the epidemic. Monteiro et al. [88] did dence in heterosexual populations is reduced include the influence of people who used drugs in through ART and prevention efforts, these models Cabo Verde, but stressed the need to strengthen do show that addressing the needs of key popula- surveillance among people using drugs to detect if tions will likely become increasingly important to injection practices start and then add injecting drug effective resource allocation in some places. use into the model. Mukandavire et al. [82 ] prepared models for Dakar, Senegal, a city with 6% prevalence in FSW and 30% Geographic risk heterogeneity is important in MSM in 2016. They found same-sex behaviors to to planning effective responses be the primary contributor to new infections in 2015, 51.4% compared with 13.8% through sex Although the preceding sections imply the contri- work. Meeting MSM’s prevention needs would bution of key populations in WCA is generally avert 64.1% of new infections over the next 10 years, greater than in ESA, such generalizations are not while strengthening sex work programs would enough to direct programs appropriately. The influ- only avert 13.6% because of existing long-term ence of key populations varies greatly between and 348 www.co-hivandaids.com Volume 14 Number 5 September 2019 Evolving HIV epidemics Brown and Peerapatanapokin within countries. McGillen et al. [89] ran dynamical Yotebieng et al. [95] and Wolf et al. stress how models for 18 high-burden SSA countries broken the lack of key population data along the treat- into 203 sub-national regions and found substantial ment cascade (prevalence, size estimates, facilita- heterogeneity in risk by different sub-populations tors and barriers) at national and sub-national and by geography. Their analysis showed focusing levels in SSA, presents a major barrier to the prevention on marginalized populations in those implementation of effective ‘treat all’ strategies countries could avert 70% more infections than a for these groups [96]. It also contributes to an strategy based on current less-focused targeting absence of modeling work addressing ‘treat all’ approaches, but the most effective combination for key populations, highlighted by Kimmel of program components varied from one sub- et al. [97]. Past reviews have stressed the serious national region to the next. In a modeling analysis deficit in cascade information for key population, of six counties in the Nyanza region of Kenya, both in SSA and globally [98,99], and efforts are Bershteyn et al. [90] found substantial geographic underway to improve this knowledge base [96]. heterogeneity in the contribution of sex work to The cascade is the next major focus for modelers as transmission, requiring different intervention identifying gaps along the cascade and estimating mixes to efficiently curtail transmission. Similarly, the effects that programs can have will impact a Goals model analysis of the NSP in Mozambique both transmission among and quality of care for found that in the North sex work interventions key populations and their partners. would have the greatest impact, whereas in the The need for this data to guide programs, Center and South, voluntary male medical circum- inform models to target them appropriately and cision would be most impactful [91]. Although scale-up programs to meet key population needs is there has recently been a substantial focus on geo- urgent. This makes the effort by the Global HIV spatial variability in prevalence and incidence in Research Group to systematize the collection of key targeting responses [92], these studies highlight population data globally and disseminate them in that it is also important to focus on sub-national easy-to-use forms a critical first step in identifying risk heterogeneity, including risk among key pop- ongoing gaps and translating key population data && ulations, to optimize prevention investments in and models into actions and accountability [13 ]. each location. However, the knowledge base at As this data comes online, modelers should use it to present is insufficient to most efficiently allocate expand the number of countries in SSA with prevention resources [10]. This increases the impor- dynamical models that include key populations tance of recent efforts to systematize this data and and improve models in countries already doing make it more readily available for program plan- so. The gaps identified must guide data collection && ning, analysis and modeling [13 ]. efforts to ensure that key populations are appropri- ately reflected in future national models and strategic planning. Data for key populations remains limited The more in-depth analytic models in Senegal, CONCLUSIONS AND IMPLICATIONS FOR Kenya and Cote d’Ivoire were each based on exten- IMPROVING HIV RESPONSES sive data collection exercises, gathering both cur- rent and historical data on prevalence, behaviors, The analysis here shows that for most ‘concentrated’ programs and size estimates in both key populations epidemics outside of sub-Saharan Africa, key pop- & & and those outside them [81 ,82 ,83]. However, ulations must remain at the core of epidemic extensive reviews of the data in West and Central responses. Although the proportion of prevalent Africa have found that whereas some data and pro- HIV infections among the ‘general population’ gram targets for sex workers were often available, grows over time, this growth is primarily reflecting much less data was available for MSM and that data turnover in key populations and transmission from for PWID was rare and incomplete [10,93]. The current and former key population members to their situation has been slowly improving over time, intimate partners. It is not that HIV is being heavily and the improvements in data have increased transmitted within the general population other UNAIDS static estimates of the proportion of new than to intimate partners of key populations, but infections among key populations and their imme- that HIV acquired while in key populations is cur- diate partners in WCA from 27% in 2014 to 40% in rently found in the ‘general population’. Only a 2017 [21,94]. Degenhardt et al. [46] document the limited amount of transmission is between mem- growth of PWID data in the countries of SSA bers of the general population with no relationship between 2007 and 2017, but few countries have to key populations. So, in most cases, concentrated translated this into programs [9]. epidemics do not go ‘generalized’. 1746-630X Copyright 2019 The Author(s). Published by Wolters Kluwer Health, Inc. www.co-hivandaids.com 349 Concentrated epidemics This has major implications for responses in and then disseminating this information widely for what we consider to be concentrated epidemics decision-making. As ART and prevention efforts today: bring down incidence outside key populations, these populations will likely play an increasing role (1) The focus in prevention efforts should remain in new and downstream infections and deaths, on key populations and not shift to general especially as they often have less access to treatment population efforts targeting lower risk popula- and prevention. Once again, the truism is wrong, tions. Resources spent on general population key populations do matter in generalized epidemics. efforts carry major opportunity costs in forgone It is time to put the concept of generalized and averted infections and deaths among the key concentrated epidemics to rest; it has outlived its populations and their intimate partners. usefulness. It encourages those not grounded in HIV (2) Programs for key populations should expand epidemiology to assume a static view of dynamic efforts and research to prevent transmission epidemics, when the drivers of the epidemic are to the intimate partners of current and former evolving over time as behaviors, the effects of pre- key population members, including clients. vention and expanded treatment programs, and Research should be undertaken to develop inno- prevalence among different populations change. vative programs to reach HIV-positive former This in turn leads to simplistic and inefficient key population members for testing, ART access choices about where to focus available resources, and partner referrals for testing and PrEP. based on an incorrect picture of the epidemic. It (3) Program resources should be allocated among is also time to bury the concept of ‘regional’ and populations in a way that maximally reduces ‘national’ epidemics. In each of the regions, while downstream infections and deaths considering there may be aggregate patterns, when one looks at each country’s unique epidemic situation. the country level, the contributions of key and non- key populations varies greatly from country to coun- In the epidemics of sub-Saharan Africa, which try, affected by the patterns of risk and the impacts most still view as generalized, the expanding data- of past prevention and treatment programs. In every base clearly demonstrates that key populations have region, the epidemics are as diverse as the histories greatly elevated prevalence compared with the pop- and cultures of the countries and they require dif- ulation at large. Current responses are not adequate ferent responses. Even within a country, epidemics or equitable and key populations continue to dis- are rarely homogeneous; the intensity and drivers of proportionately feel the impacts of HIV. Moreover, the epidemic vary from place to place and that calls static analyses, such as the Modes of Transmission for varying responses. Expanded work to identify model underestimate the influence of key popula- sub-national risk patterns and assess local key driv- tions on epidemics by not accounting for down- ers is needed to adjust the response in different areas stream infections, making mobilization of the to maximize effectiveness. This may require devel- necessary responses difficult. In the case of sex work, oping simplified models or easy to apply algorithms modeling has shown the magnitude of downstream that are more accessible to local health authorities. contributions can be large, especially in early stages Instead, these simplistic conceptions of epidem- of generalized epidemics. Even today, expanding ics should be replaced by the recognition that ‘each program coverage for sex workers, proven effective epidemic is unique’. This should be engendered in a in reducing transmission, will pay substantial bene- systematic approach focused on understanding fits at comparatively low cost in most countries. Yet, local variations in risk and prevalence between coverage in many countries remains far from uni- and within countries and between and within versal [11]. sub-populations. Starting by identifying and filling Too little attention has been paid to the needs of important data gaps for all local populations, coun- MSM and PWID in sub-Saharan Africa, two popula- tries should then apply dynamical models to appro- tions with even higher prevalence compared with priately target and adapt responses as epidemics the population at large than sex workers. This has evolve. These models must include all important been aided and abetted by a lack of reliable data on populations, including key populations, even in prevalence, size estimates and risk behaviors and countries where people think them unimportant. serious stigma and discrimination. Although data One of the lessons learned in the Global Estimates are improving in recent years and efforts to system- process is that building something into models atize and make this data accessible are underway drives the data generation process, and until the && [13 ], the next major challenge will be using this data weaknesses for key populations are addressed, data to construct models accurately reflecting the either by improving data systems or using expanded overall impact these groups have on local epidemics program data, their role in epidemics and their 350 www.co-hivandaids.com Volume 14 Number 5 September 2019 Evolving HIV epidemics Brown and Peerapatanapokin programmatic needs will not be properly assessed REFERENCES AND RECOMMENDED and met. Accordingly, all relevant key populations READING Papers of particular interest, published within the annual period of review, have should be included in future modeling work regard- been highlighted as: less of region or country, even if informed assump- of special interest && of outstanding interest tions need to be made in the short-term. With more comprehensive data and models, resource alloca- 1. UNAIDS/WHO Working Group on Global HIV/AIDS and STI Surveillance. Guidelines for second generation HIV surveillance. Geneva: UNAIDS/WHO; tions can be equitably adapted for maximum impact as epidemics evolve over time. 2. UNAIDS. Practical guidelines for intensifying HIV prevention: towards universal access. Geneva: Joint United National Programme on HIV/AIDS; Finally, improving data and building detailed 2007 . models explaining local epidemiological realities 3. Piot P, Coll Seck AM. International response to the HIV/AIDS epidemic: planning for success. Bull World Health Organ 2001; 79:1106 – 1112. does no good if it does not drive responses. These 4. Stover J, Walker N, Garnett GP, et al. Can we reverse the HIV/AIDS pandemic efforts must be coupled with vigorous and targeted with an expanded response? Lancet 2002; 360:73 – 77. 5. Ainsworth M, Teokul W. Breaking the silence: setting realistic priorities for dissemination plans to convey the uniqueness of the AIDS control in less-developed countries. Lancet 2000; 356:55 – 60. local epidemic to key programmatic policy planners, 6. UNAIDS. Speech by Michel Sidibe at the Sixth International Conference on HIV and AIDS in Eastern Europe and Central Asia, Moscow, Russia 18 communities and budget authorities in a clear and April 2018. Geneva: UNAIDS; 2018. unambiguous manner. It is critical to counter the 7. Human Rights Watch. Fueling the Philippines’ HIV epidemic. Government barriers to condom use by men who have sex with men. USA: Human Rights deeply ingrained generalized/concentrated para- Watch; 2016. digm at the highest levels, which so often misdirects 8. Kumaranayake L, Watts C. Resource allocation and priority setting of HIV/ AIDS interventions: addressing the generalized epidemic in sub-Saharan prevention, fails to address key population needs, Africa. Int J Dev 2001; 13:451 – 466. keeps key population resources constrained and 9. Larney S, Peacock A, Leung J, et al. Global, regional, and country-level coverage of interventions to prevent and manage HIV and hepatitis C among gives stigma and discrimination additional breath- people who inject drugs: a systematic review. Lancet Glob Health 2017; ing room. Regardless of region, the data and models 5:e1208 – e1220. 10. MacAllister J, Sherwood J, Galjour J, et al. A comprehensive review of demonstrate the urgent need to refocus local available epidemiologic and HIV service data for female sex workers, men responses on the populations at the greatest risk if who have sex with men, and people who inject drugs in select West and Central African countries. J Acquir Immune Defic Syndr 2015; 68 Suppl AIDS is to end by 2030. 2:S83 – S90. 11. Shannon K, Crago AL, Baral SD, et al. The global response and unmet actions for HIV and sex workers. Lancet 2018; 392:698 – 710. Acknowledgements 12. Beyrer C, Baral SD, Collins C, et al. The global response to HIV in men who The authors gratefully acknowledge the hard work and have sex with men. Lancet 2016; 388:198 – 206. 13. Rao A, Schwartz S, Sabin K, et al., Global. HIV Research Group. HIV-related support of the national AEM country teams in && data among key populations to inform evidence-based responses: protocol of Bangladesh, Cambodia, Indonesia, Lao PDR, Malaysia, a systematic review. Syst Rev 2018; 7:220. Introduces a critically important effort to collate and systematize key population Myanmar, Nepal, Pakistan, the Philippines, Thailand data globally to make it more available for modeling, analysis and decision-making. and Viet Nam whose AEM-based national models for 14. Brown T, Peerapatanapokin W. The Asian Epidemic Model: a process model for exploring HIV policy and programme alternatives in Asia. Sex Transm Infect 2017 were essential inputs informing this article. Thanks 2004; 80 Suppl 1:i19 – i24. are also due to Dr Swarup Sarkar and the World Health 15. Baggaley RF, Owen BN, Silhol R, et al. Does per-act HIV-1 transmission risk through anal sex vary by gender? An updated systematic review and meta- Organization South-East Asia Regional Office (WHO analysis. Am J Reprod Immunol 2018; 80:e13039. SEARO) who organized the think tank meeting on ‘Revis- 16. Boily MC, Baggaley RF, Wang L, et al. Heterosexual risk of HIV-1 infection per sexual act: systematic review and meta-analysis of observational studies. iting strategies for intervention among key populations’ Lancet Infect Dis 2009; 9:118 – 129. in February 2018 that eventually led to this article. 17. Fazito E, Cuchi P, Mahy M, Brown T. Analysis of duration of risk behaviour for key populations: a literature review. Sex Transm Infect 2012; 88 Suppl 2:i24 – i32. Financial support and sponsorship 18. Weniger BG, Limpakarnjanarat K, Ungchusak K, et al. The epidemiology of HIV infection and AIDS in Thailand. AIDS 1991; 5 Suppl 2:S71 – 85. The authors wish to thank WHO SEARO for their support 19. Commission on, AIDS in Asia. Redefining AIDS in Asia, crafting an effective response. Report of the Commission on AIDS in Asia. New Delhi: Oxford of the recent scenario work for Asian countries used in University Press; 2008. this article. The authors further offer their deep gratitude 20. Saidel T, Des Jarlais DC, Peerapatanapokin W, et al. Potential impact of HIV among IDUs on heterosexual transmission in Asian settings: scenarios from to the Global Fund to Fight AIDS, Tuberculosis and the Asian Epidemic Model. Int J Drug Policy 2003; 14:63 – 74. Malaria; the Gates Foundation through Avenir Health; 21. UNAIDS. Miles to go: closing gaps, breaking barriers, righting injustices. Geneva: Joint United Nations Programme on HIV/AIDS; 2018. the Joint United Nations Programme on HIV/AIDS; 22. Luz PM, Osher B, Grinsztejn B, et al. The cost-effectiveness of HIV Family Health International, the United States Agency preexposure prophylaxis in men who have sex with men and transgender womenat highriskofHIV infectioninBrazil. JInt AIDS Soc2018; for International Development and the World Bank who 21:e25096. have supported the ongoing development and application 23. Broughton EI, Nunez O, Arana R, Oviedo A. Effectiveness and efficiency of improving HIV service provision for key populations in Nicaragua. Front Public of the AIDS Epidemic Model at national level over the last Health 2016; 4:249. two decades. Their support made the national models 24. Monteiro JF, Marshall BD, Escudero D, et al. Preventing HIV transmission among partners of HIV-positive male sex workers in Mexico City: a modeling used here possible. study. AIDS Behav 2015; 19:1579 – 1588. 25. Caro-Vega Y, del Rio C, Lima VD, et al. Estimating the impact of earlier ART initiation and increased testing coverage on HIV transmission among men Conflicts of interest who have sex with men in Mexico using a mathematical model. PLoS One There are no conflicts of interest. 2015; 10:e0136534. 1746-630X Copyright 2019 The Author(s). Published by Wolters Kluwer Health, Inc. www.co-hivandaids.com 351 Concentrated epidemics 26. Goodreau SM, Carnegie NB, Vittinghoff E, et al. What drives the US and 51. Okal J, Raymond HF, Tun W, et al. Lessons learned from respondent-driven Peruvian HIV epidemics in men who have sex with men (MSM)? PLoS One sampling recruitment in Nairobi: experiences from the field. BMC Res Notes 2012; 7:e50522. 2016; 9:158. 27. Gomez GB, Borquez A, Caceres CF, et al. The potential impact of preexpo- 52. Quaye S, Fisher Raymond H, Atuahene K, et al. Critique and lessons learned sure prophylaxis for HIV prevention among men who have sex with men and from using multiple methods to estimate population size of men who have sex transwomen in Lima, Peru: a mathematical modelling study. PLoS Med 2012; with men in Ghana. AIDS Behav 2015; 19 Suppl 1:S16 – 23. 9:e1001323. 53. Vuylsteke B, Sika L, Semde G, et al. Estimating the number of female sex 28. Borquez A, Guanira JV, Revill P, et al. The impact and cost-effectiveness of workers in Cote d’Ivoire: results and lessons learned. Trop Med Int Health combined HIV prevention scenarios among transgender women sex-workers 2017; 22:1112 – 1118. in Lima, Peru: a mathematical modelling study. Lancet Public Health 2019; 54. Datta A, Lin WW, Rao A, et al. Bayesian estimation of MSM population size in 4:e127 – e136. Cote d’Ivoire. Stat Public Policy 2019; 6:1 – 13. 29. Borquez A, Beletsky L, Nosyk B, et al. The effect of public health-oriented 55. Wesson PD, Adhikary R, Jonas A, et al. Estimating the population size of drug law reform on HIV incidence in people who inject drugs in Tijuana, female sex workers in Namibia using a respondent-driven sampling adjust- Mexico: an epidemic modelling study. Lancet Public Health 2018; ment to the reverse tracking method: a novel approach. JMIR Public Health 3:e429 – e437. Surveill 2019; 5:e11737. 30. Burgos JL, Patterson TL, Graff-Zivin JS, et al. Cost-effectiveness of combined 56. Weir SS, Baral SD, Edwards JK, et al. Opportunities for enhanced strategic sexual and injection risk reduction interventions among female sex workers use of surveys, medical records, and program data for HIV surveillance of key who inject drugs in two very distinct Mexican Border cities. PLoS One 2016; populations: scoping review. JMIR Public Health Surveill 2018; 4:e28. 11:e0147719. 57. Kavanagh MM, Baral SD, Milanga M, Sugarman J. Biometrics and public 31. Wirtz AL, Pretorius C, Beyrer C, et al. Epidemic impacts of a community health surveillance in criminalised and key populations: policy, ethics, and empowerment intervention for HIV prevention among female sex workers in human rights considerations. Lancet HIV 2019; 6:e51 – e59. generalized and concentrated epidemics. PLoS One 2014; 9:e88047. 58. Davis SL, Goedel WC, Emerson J, Guven BS. Punitive laws, key population 32. Mabileau G, Scutelniciuc O, Tsereteli M, et al. Intervention packages to size estimates, and Global AIDS Response Progress Reports: an ecological reduce the impact of HIV and HCV infections among people who inject study of 154 countries. J Int AIDS Soc 2017; 20:21386. drugs in eastern europe and central asia: a modeling and cost-effectiveness 59. KELIN, the Kenya Key Populations Consortium. Everyone said no’: Biometrics, study. Open Forum Infect Dis 2018; 5:ofy040. HIV and Human Rights, a Kenya Case Study. Nairobi: KELIN; 2018. 33. Cepeda JA, Eritsyan K, Vickerman P, et al. Potential impact of implementing 60. Vandepitte J, Lyerla R, Dallabetta G, et al. Estimates of the number of female and scaling up harm reduction and antiretroviral therapy on HIV prevalence sex workers in different regions of the world. Sex Transm Infect 2006; 82 and mortality and overdose deaths among people who inject drugs in two Suppl 3:iii18 – iii25. Russian cities: a modelling study. Lancet HIV 2018; 5:e578 – e587. 61. Sabin K, Zhao J, Garcia Calleja JM, et al. Availability and quality of size 34. Idrisov B, Murphy SM, Morrill T, et al. Implementation of methadone therapy for estimations of female sex workers, men who have sex with men, people who opioid use disorder in Russia - a modeled cost-effectiveness analysis. Subst inject drugs and transgender women in low- and middle-income countries. Abuse Treat Prev Policy 2017; 12:4. PLoS One 2016; 11:e0155150. 35. Vickerman P, Platt L, Jolley E, et al. Controlling HIV among people who inject 62. Diamond M. Homosexuality and bisexuality in different populations. Arch Sex drugs in Eastern Europe and Central Asia: insights from modeling. Int J Drug Behav 1993; 22:291 – 310. Policy 2014; 25:1163 – 1173. 63. Marcus U, Hickson F, Weatherburn P, et al. Estimating the size of the MSM 36. Mumtaz GR, Awad SF, Feizzadeh A, et al. HIV incidence among people who populations for 38 European countries by calculating the survey-surveillance inject drugs in the Middle East and North Africa: mathematical modelling discrepancies (SSD) between self-reported new HIV diagnoses from the analysis. J Int AIDS Soc 2018; 21:e25102. European MSM internet survey (EMIS) and surveillance-reported HIV diag- 37. Kouyoumjian SP, El Rhilani H, Latifi A, et al. Mapping of new HIV infections in noses among MSM in. BMC Public Health 2013; 13:919. Morocco and impact of select interventions. Int J Infect Dis 2018; 68:4 – 12. 64. Baral S, Turner RM, Lyons CE, et al. Population size estimation of gay and 38. Mumtaz GR, Kouyoumjian SP, Hilmi N, et al. The distribution of new HIV & bisexual men and other men who have sex with men using social media-based infections by mode of exposure in Morocco. Sex Transm Infect 2013; 89 platforms. JMIR Public Health Surveill 2018; 4:e15. Suppl 3:iii49 – iii56. An example of an innovative approach to size estimation for MSM using social 39. Stuart RM, Grobicki L, Haghparast-Bidgoli H, et al. How should HIV resources media to challenge the traditional under-estimation of MSM. && be allocated? Lessons learnt from applying Optima HIV in 23 countries. J Int 65. International Lesbian, Gay, Bisexual, Trans and Intersex Association. In: AIDS Soc 2018; 21:e25097. Mendos L. State-Sponsored Homophobia 2019. Geneva: International Les- Highlights important lessons learned in doing optimization with dynamic models bian, Gay, Bisexual, Trans and Intersex Association; 2019. including all key populations in a variety of epidemic settings. 66. Gouws E, Cuchi P. International collaboration on estimating HIV incidence by 40. Kelly SL, Martin-Hughes R, Stuart RM, et al. The global Optima HIV allocative modes of transmission. Focusing the HIV response through estimating the efficiency model: targeting resources in efforts to end AIDS. Lancet HIV major modes of HIV transmission: a multicountry analysis. Sex Transm Infect 2018; 5:e190 – e198. 2012; 88 Suppl 2:i76 – i85. 41. Scorgie F, Chersich MF, Ntaganira I, et al. Socio-demographic characteristics 67. Colvin M, Gorgens-Albino M, Kasedde S. Analysis of HIV Prevention Re- and behavioral risk factors of female sex workers in sub-saharan Africa: a sponse and Modes of HIV Transmission: The UNAIDS-GAMET Supported systematic review. AIDS Behav 2012; 16:920 – 933. Synthesis Process. Johannesburg: UNAIDS Regional Support Team Eastern 42. Stoebenau K, Heise L, Wamoyi J, Bobrova N. Revisiting the understanding of and Southern Africa; 2008. ‘transactional sex’ in sub-Saharan Africa: A review and synthesis of the 68. UNAIDS and the World Bank. New HIV Infections by mode of transmission in literature. Soc Sci Med 2016; 168:186 – 197. West Africa: A Multi-Country Analysis. Dakar, Senegal: UNAIDS Regional 43. Wamoyi J, Ranganathan M, Kyegombe N, Stoebenau K. Improving the Support Team for West and Central Africa; 2010. measurement of transactional sex in Sub-Saharan Africa: a critical review. 69. Global HIV/AIDS Program. West Africa HIV/AIDS epidemiology and re- J Acquir Immune Defic Syndr 2019; 80:367 – 374. sponse synthesis: implications for prevention. Washington DC.: World Bank; 44. Friedland BA, Sprague L, Nyblade L, et al. Measuring intersecting stigma 2008. among key populations living with HIV: implementing the people living with 70. USAID AWARE II Project Team. The Epidemiology of HIV epidemics in the HIV Stigma Index 2.0. J Int AIDS Soc 2018; 21 Suppl 5:e25131. 21-country West Africa Region: the impact of most at risk populations 45. Stahlman S, Sanchez TH, Sullivan PS, et al. The prevalence of sexual behavior (MARPs). 2011. stigma affecting gay men and other men who have sex with men across Sub- 71. Boily MC, Pickles M, Alary M, et al. What really is a concentrated HIV epidemic Saharan Africa and in the United States. JMIR Public Health Surveill 2016; and what does it mean for West and Central Africa? Insights from mathe- 2:e35. matical modeling. J Acquir Immune Defic Syndr 2015; 68 Suppl 2:S74 – S82. 46. Degenhardt L, Peacock A, Colledge S, et al. Global prevalence of injecting 72. Wilson D. HIV epidemiology: a review of recent trends and lessons. World drug use and sociodemographic characteristics and prevalence of HIV, HBV, Bank. 2006. and HCV in people who inject drugs: a multistage systematic review. Lancet 73. Mishra S, Pickles M, Blanchard JF, et al. Validation of the modes of transmis- Global health 2017; 5:e1192 – e1207. sion model as a tool to prioritize HIV prevention targets: a comparative 47. UNAIDS. UNAIDS Data 2018. Geneva, Switzerland: Joint United National modelling analysis. PLoS One 2014; 9:e101690. Programme on HIV/AIDS; 2018. 74. Shubber Z, Mishra S, Vesga JF, Boily MC. The HIV Modes of Transmission 48. Abdul-Quader AS, Baughman AL, Hladik W. Estimating the size of key model: a systematic review of its findings and adherence to guidelines. J Int populations: current status and future possibilities. Curr Opin HIV AIDS AIDS Soc 2014; 17:18928. 2014; 9:107 – 114. 75. Mishra S, Pickles M, Blanchard JF, et al. Distinguishing sources of HIV 49. Okiria AG, Bolo A, Achut V, et al. Novel approaches for estimating female sex transmission from the distribution of newly acquired HIV infections: why is worker population size in conflict-affected South Sudan. JMIR Public Health it important for HIV prevention planning? Sex Transm Infect 2014; 90:19 – 25. Surveill 2019; 5:e11576. 76. Mishra S, Boily MC, Schwartz S, et al. Data and methods to characterize the 50. Apodaca K, Doshi RH, Ogwal M, et al. Capture-recapture among men who role of sex work and to inform sex work programs in generalized HIV have sex with men and among female sex workers in 11 towns in Uganda. epidemics: evidence to challenge assumptions. Ann Epidemiol 2016; JMIR Public Health Surveill 2019; 5:e12316. 26:557 – 569. 352 www.co-hivandaids.com Volume 14 Number 5 September 2019 Evolving HIV epidemics Brown and Peerapatanapokin 77. Low A, Nagot N, Konate I, et al. Potential impact of existing interventions and 86. Volz EM, Ndembi N, Nowak R, et al. Phylodynamic analysis to inform pre- of antiretroviral use in female sex workers on transmission of HIV in Burkina && vention efforts in mixed HIV epidemics. Virus Evol 2017; 3:vex014. Faso: a modeling study. J Acquir Immune Defic Syndr 2015; 68 Suppl Demonstrates how phylodynamic modeling work can be used to assess transmis- 2:S180 – 188. sion between MSM and female partners. 78. Steen R, Hontelez JA, Veraart A, et al. Looking upstream to prevent HIV 87. Rhodes T, Guise A, Ndimbii J, et al. Is the promise of methadone Kenya’s transmission: can interventions with sex workers alter the course of HIV solution to managing HIV and addiction? A mixed-method mathematical epidemics in Africa as they did in Asia? AIDS 2014; 28:891 – 899. modelling and qualitative study. BMJ Open 2015; 5:e007198. 79. Williams JR, Alary M, Lowndes CM, et al. Positive impact of increases in 88. Monteiro JF, Galea S, Flanigan T, et al. Evaluating HIV prevention strategies condom use among female sex workers and clients in a medium HIV pre- for populations in key affected groups: the example of Cabo Verde. Int J Public valence epidemic: modelling results from Project SIDA1/2/3 in Cotonou, Health 2015; 60:457 – 466. Benin. PLoS One 2014; 9:e102643. 89. McGillen JB, Anderson SJ, Dybul MR, Hallett TB. Optimum resource alloca- 80. Kahn JG, Bollinger LA, Stover J, Marseille E. Improving the efficiency of the tion to reduce HIV incidence across sub-Saharan Africa: a mathematical && HIV/AIDS policy response: a guide to resource allocation modeling. In: modelling study. Lancet HIV 2016; 3:e441 – e448. Holmes KK, Bertozzi S, editors. Major infectious diseases. Washington, D.C.: 90. Bershteyn A, Mutai KK, Akullian AN, et al. The influence of mobility among World Bank; 2017. high-risk populations on HIV transmission in Western Kenya. Infect Dis Model A good overview of the primary dynamic models being used for resource allocation 2018; 3:97 – 106. and optimization with support from UNAIDS, Global Fund and World Bank. 91. Korenromp EL, Gobet B, Fazito E, et al. Impact and cost of the HIV/AIDS 81. Maheu-Giroux M, Vesga JF, Diabate S, et al. Changing dynamics of HIV national strategic plan for Mozambique, 2015-2019–projections with the transmission in Cote d’Ivoire: modeling who acquired and transmitted infec- Spectrum/Goals Model. PLoS One 2015; 10:e0142908. tions and estimating the impact of past HIV interventions. J Acquir Immune 92. Meyer-Rath G, McGillen JB, Cuadros DF, et al. Targeting the right interven- Defic Syndr 2017; 75:517 – 527. tions to the right people and places: the role of geospatial analysis in HIV A good example of how the contribution of sex work to a sub-Saharan African program planning. AIDS 2018; 32:957 – 963. epidemic can change substantially over the course of the epidemic. 93. Djomand G, Quaye S, Sullivan PS. HIV epidemic among key populations in 82. Mukandavire C, Walker J, Schwartz S, et al. Estimating the contribution of key west Africa. Curr Opin HIV AIDS 2014; 9:506 – 513. & populations towards the spread of HIV in Dakar, Senegal. J Int AIDS Soc 94. UNAIDS. AIDS by the numbers. Geneva: UNAIDS; 2016. 2018; 21(Suppl 5):e25126. 95. Yotebieng M, Brazier E, Addison D, et al. Research priorities to inform ‘Treat An example of how MSM in some settings in sub-Saharan Africa may be con- All’ policy implementation for people living with HIV in sub-Saharan Africa: a tributing more to new infections than other populations when downstream trans- consensus statement from the International epidemiology Databases to mission is taken into account. Evaluate AIDS (IeDEA). J Int AIDS Soc 2019; 22:e25218. 83. Cremin I, McKinnon L, Kimani J, et al. PrEP for key populations in combination 96. Lillie TA, Baer J, Adams D, et al. Think global, act local: the experience of HIV prevention in Nairobi: a mathematical modelling study. Lancet HIV 2017; Global Fund and PEPFAR joint cascade assessments to harmonize and 4:e214 – e222. strengthen key population HIV programmes in eight countries. J Int AIDS Soc 84. Maheu-Giroux M, Vesga JF, Diabate S, et al. Population-level impact of an 2018; 21 Suppl 5:e25125. accelerated HIV response plan to reach the UNAIDS 90-90-90 target in Cote 97. Kimmel AD, Bono RS, Keiser O, et al. Mathematical modelling to inform ’treat d’Ivoire: insights from mathematical modeling. PLoS Med 2017; all’ implementation in sub-Saharan Africa: a scoping review. J Virus Erad 14:e1002321. 2018; 4(Suppl):47 – 54. 85. Maheu-Giroux M, Diabate S, Boily MC, et al. Cost-effectiveness of acceler- 98. Lancaster KE, Cernigliaro D, Zulliger R, Fleming PF. HIV care and treatment && ated HIV response scenarios in Cote d’Ivoire. J Acquir Immune Defic Syndr experiences among female sex workers living with HIV in sub-Saharan Africa: 2019; 80:503 – 512. a systematic review. Afr J AIDS Res 2016; 15:377 – 386. A good example of the improved cost-effectiveness of focusing on key populations 99. Risher K, Mayer KH, Beyrer C. HIV treatment cascade in MSM, people who in generalized settings when facing cost constraints. inject drugs, and sex workers. Curr Opin HIV AIDS 2015; 10:420 – 429. 1746-630X Copyright 2019 The Author(s). Published by Wolters Kluwer Health, Inc. www.co-hivandaids.com 353
Current Opinion in HIV and Aids – Wolters Kluwer Health
Published: Sep 1, 2019
You can share this free article with as many people as you like with the url below! We hope you enjoy this feature!
Read and print from thousands of top scholarly journals.
Already have an account? Log in
Bookmark this article. You can see your Bookmarks on your DeepDyve Library.
To save an article, log in first, or sign up for a DeepDyve account if you don’t already have one.
Copy and paste the desired citation format or use the link below to download a file formatted for EndNote
Access the full text.
Sign up today, get DeepDyve free for 14 days.
All DeepDyve websites use cookies to improve your online experience. They were placed on your computer when you launched this website. You can change your cookie settings through your browser.